This is a resubmission for a 25th year of a project investigating how working memory (WM), the small amount held in mind to carry out cognitive tasks, increases with development in the elementary school years and to adulthood. Much of the proposal has been rewritten with reviewers' comments taken to heart. They lauded the past progress; it helps establish a key component of the human mind, an attention-related, central component of working memory (WM) limited to a few meaningful units, or chunks. The reviewers, however, highlighted 2 general problems with the previously proposal. (1) There was insufficient theoretical rationale for the studies that were proposed, with an insufficiently clear new direction for the research, and (2) the research approach was perceived as too incremental and not theoretically bold enough, not breaking enough new ground beyond past progress. In response, the proposed research now contributes to one central question: What are the mechanisms of the developmental increase in WM capacity? Although Miller (1956) discussed capacity limits in terms of the number of meaningful chunks recalled, researchers since then have mostly avoided this topic because of the great difficulty of measuring chunks and distinguishing capacity from different components of WM. Yet, chunk capacity is critical to understand as it may determine the permissible complexity of human thought. Isolating chunk capacity from other processes has been a striking and difficult achievement, which has gained popularity (e.g., the review by Cowan, 2001, cited over 1,800 times, pointing to a constant-capacity mechanism). Our developmental investigations so far have proven especially important in establishing the notion that there is a core WM capacity that is separate from other mnemonic processes (e.g., the ability to combine items to form larger chunks, exclude distracting information, rehearse items verbally, or refresh them at a rapid rate). Careful developmental investigation can highlight the improvement in one skill with others held constant. What has not been accomplished yet is, first, to establish the generality of the findings. So far we have examined capacity with selective attention and rehearsal held constant only for visual array probe recognition, whereas we have examined capacity with chunk formation held constant only for verbal free recall. The first order of business is therefore to determine whether the WM faculties that develop during the elementary school years include only domain-specific stores, or also a central, attention-related, core WM faculty observable in tests of recognition and recall, i both verbal and visual materials. Developmental investigation can help us to understand how a core capacity limit is dependent on attention, whether items are held concurrently or refreshed sequentially, and whether a core WM is based on retention of a certain number of items or features. We incorporate behavioral and neuroimaging methods and math modeling to understand normal WM development, as it guides our understanding of (1) cognitive development and (2) how we might treat cognitive deficits, i.e., which basic skills might need training or other remediation.